How useful is the DSM-5 severity indicator in bulimia nervosa? A clinical study including a measure of impairment.

Abstract

The severity criterion used in DSM-5 for bulimia nervosa (BN) was investigated in 214 individuals referred for treatment at a regional eating disorders service in the UK. In addition to comparing eating disorder symptoms, impairment secondary to these symptoms was also assessed. According to guidance in DSM-5, 94 individuals were classified as mild (43.9%), 70 as moderate (32.7%), 32 as severe (15.0%), and 8 as extreme (3.7%) levels of BN severity. Due to small numbers in the latter two groups, it was necessary to combine these to form one 'severe/extreme' group. Analyses on these three groups suggested no group effect on demographic variables but differences were seen on measures of eating pathology, psychological distress, and psychosocial impairment between the mild group and other groups. Individuals in the moderate and severe/extreme groups scored comparably on most measures of pathology and impairment. The results are broadly consistent with past studies on community samples although together question the demarcation between moderate and more severe groups of individuals with BN.